[R] Best way to predict

Peter Langfelder peter.langfelder at gmail.com
Thu May 31 05:24:40 CEST 2012


On Wed, May 30, 2012 at 4:07 PM, Chris Burns <chris.burns2 at gmail.com> wrote:
> I have a huge matrix of unspecified covariates and the corresponding sales
> for them.  What is the best way to predict the sales from the covariates?
>

Don't want to sound rude, but given your very vague problem
specifications, the best way seems to be to consult a local
statistician or machine learning specialist. You can also look up
literature on prediction or machine learning, in particular predictors
like k nearest neighbors, Random Forest, Support Vector Machines etc.
Once you decide which predictor(s) to test and use, finding a good R
implementation will be trivial.

HTH,

Peter



More information about the R-help mailing list